Behavioral Outcomes of Parent-Child Interaction

Therapy
and Triple PPositive Parenting Program: A
Review and
Meta-Analysis
Rae Thomas Melanie J. Zimmer-Gembeck (2007)
Abstract We conducted a review and meta-analyses of 24 studies to evaluate and
compare the outcomes of two widely disseminated parenting interventionsParent-Child
Interaction
Therapy and Triple P-Positive Parenting Program. Participants
in all studies were caregivers and 3- to 12-year-old
children. In general, our analyses revealed positive effects
of both interventions, but effects varied depending on intervention
length, components, and source of outcome data.
Both interventions reduced parent-reported child behavior
and parenting problems. The effect sizes for PCIT were
large when outcomes of child and parent behaviors were
assessed with parent-report, with the exclusion of Abbreviated
PCIT, which had moderate effect sizes. All forms of
Triple P had moderate to large effects when outcomes were
parent-reported child behaviors and parenting, with the exception
of Media Triple P, which had small effects. PCIT
and an enhanced version of Triple P were associated with
improvements in observed child behaviors. These findings
provide information about the relative efficacy of two programs
that have received substantial funding in the USA and
Australia, and findings should assist in making decisions
about allocations of funding and dissemination of these parenting
interventions in the future.
Behavioral parent training interventions are widely used because
the evidence shows they are efficacious (Serketich &
Dumas, 1996). Although parenting interventions focusing on
child behavior management are widespread there are a variety
of theoretical foundations and delivery formats for these
interventions (Kumpfer & Alvarado, 2003). Nevertheless,
behavioral parent training, where parents participate in skills
training, is included in many interventions with demonstrated
efficacy (Serketich & Dumas, 1996). Often, behavioral parent
training involves discussing differential reinforcement
and timeout procedures, and an emphasis is placed on the
role the parent plays in the development and maintenance of
child problem behaviors (Hollenstein, Granic, Stoolmiller,
& Snyder, 2004). Yet, interventions might only include parents,
might involve family skills training, where the parents
and children are taught skills and have time to practice these
together with a therapist, or interventions may involve family
therapy in which the family unit receives intervention with
less of the content directed at skill development.
Clinical benefits of behavioral parent training interventions
were described in one meta-analysis of 26 published
studies (Sereketich & Dumas, 1996). In another systematic
review (Barlow & Stewart-Brown, 2000), 16 randomized

controlled trials of group parenting interventions were determined

to be efficacious for improving child behavior when
compared to waitlists or other comparison groups. In this
review, however, the heterogeneity of the interventions and
study designs, and the difficulties in generalizing the findings
were emphasized. There have been few reviews that have
systematically compared the efficacy or effectiveness of parenting
interventions that have different primary methodologies.
Hence, the evidence of efficacy of parent interventions
pertains to the broad category of behavioral parent training
rather than particular intervention types. In past reviews,
summaries included one or two published controlled trials
to represent each of a range of studies that used differing
training methodologies.
Recently, recommendations were made regarding the
principle characteristics of effective child and family interventions
(Kumpfer & Alvarado, 2003; Nation et al., 2003).
Kumpfer and Alvarado differentiated intervention methodologies
with respects to service delivery, child only interventions,
parent only interventions, and family-focused interventions.
Family-focused prevention interventions were
cited as, in general, producing moderate to very large effect
sizes, whereas child only interventions often produced
very small effects. Although this paper was not a systematic
review of interventions, the distinction between intervention
methodologies and their differing effect on family
functioning highlights the importance of examining the efficacy
of specific types of parenting interventions, and the
potential importance of such findings for governmental policy
and decision-making with respect to dissemination of
interventions.
Parent-Child Interaction Therapy (PCIT) and Triple
PPositive Parenting Program (Triple P) are two parenting
interventions described as behavioral parent training interventions,
but PCIT and Triple P rely on different delivery
formats, and Triple P also uses a range of delivery formats,
including using the media, individual, and group intervention.
The current reviewsummarizes what is known about the
efficacy of PCIT and Triple P, and their subtypes, for three
primary reasons. First, in a previous review, an emphasis
was placed upon describing and attending to actual intervention
content and delivery formats in future meta-analyses
(Geeraert, Van den Noortgate, Grietens, & Onghena, 2004).
Hence, we described PCIT and Triple P content and delivery
formats here, and used meta-analytic techniques to assess
what is currently known about the efficacy of these two interventions,
which are both founded in social learning theory,
but have different modes of service delivery. Although using
meta-analysis to summarize the findings from trials of PCIT
and trials of Triple P is not a direct comparison of delivery
format, as other differences between the two types of
interventions and each trial were not controlled, we elected
to analyze trials of PCIT and Triple P because of the clear
differences in delivery formats and the potential importance
of such information for future research and intervention implementation.
Second, the number of trials of PCIT and Triple P has
reached a level that makes them amenable to summarization
separate from other parenting interventions. We also
expected that these findings would be of importance to families,

communities, and policymakers, because these two interventions

have been (and are continuing to be) widely disseminated
and funded in both the USA and Australia. PCIT
and Triple P have become popular in Australia in recent
years, and both have been used widely or are being disseminated
in the USA. Triple P has seen widespread growth in
Australia, with recent dissemination to the USA (Prinz &
Sanders, 2004). PCIT is used widely in the USA, with more
recent dissemination to Australia (Nixon, 2001). PCIT and
Triple P have attracted government funding in both countries
to support their implementation.
Conclusions about the effectiveness of PCIT and Triple
P for improving parenting behaviors and reducing child behavior
problems have been made in some previous literature
reviews. In one review, PCIT was described as a probably
efficacious treatment for children with externalizing and antisocial
behavior (Brestan & Eyberg, 1998), whereas Nixon
(2002) cited both PCIT and Triple P as parent training interventions
with demonstrated efficacy based upon numerous
studies with rigorous methodologies. Webster-Stratton
and Taylor (2001) cited both intervention types as empirically
supported for the reduction of child antisocial behavior.
However, a third reason for conducting the meta-analyses reported
here was that none of these reviews were systematic
reviews of all published evaluation studies of both PCIT and
Triple P, and there was no use of quantitative meta-analytic
methods to summarize findings from all studies. Based on
the conclusions of these narrative reviews and others (e.g.,
Foote, Schuhmann, Jones, & Eyberg, 1998; Sanders, Cann,
& Markie-Dadds, 2003) and the expectation that services
provide interventions that are evidence-based, Triple P
and PCIT have attracted significant funding in the USA and
Australia. As such, the evidence of efficacy and impact of
differing methodology of both Triple P and PCIT warranted
a more detailed examination. In the current study, we completed
a review and meta-analyses of all trials of Triple P and
PCIT dated between 1980 and 2004.
Theoretical foundations of PCIT and Triple P
PCIT and Triple P have been derived from social and developmental
theories. First, some PCIT and Triple P intervention
components have a foundation in the propositions of
social learning theory (Foote, Eyberg, & Schuhmann, 1998;
Sanders et al., 2003). For example, a social learning framework
has been employed to direct attention to the interactions
between family members as the source of difficulty, rather
than implying that the child or the parent is independently
responsible for all problems. For minor child misbehavior
differential reinforcement is the primary positive parenting
strategy taught in both interventions. Differential reinforcement
is a planned behavior management strategy of positively
rewarding a childs prosocial behavior whilst providing
minimal attention to a childs inappropriate behavior. In
addition to differential reinforcement, both PCIT and Triple
P focus on appropriate consequences for child misbehavior.
Positive reinforcement of the parent during interactions
with the child is also used in PCIT, but this is not directly
integrated into Triple P.
Second, applied behavior analysis, developmental models
of social competence, and developmental psychopathology

are theories described as foundations for Triple P (Sanders

et al., 2003), and attachment theory has been described as a
foundation of PCIT (Herschell, Calzada, Eyberg, &McNeil,
2002). Although PCIT and Triple P share theoretical underpinnings,
there are often differences in intervention length,
service delivery components, and how the intervention format
met their specified goals.
Triple PPositive Parenting Program
Triple P was designed to promote positive parenting and
caring relationships between parents and children aged between
2 and 16 years (Sanders et al., 2003). Triple P
has been described as a Behavioral Family Intervention
(e.g., Sanders & McFarland, 2000) with a multi-tiered
continuum of service intervention (Sanders et al., 2003).
Families are offered information on parenting and behavior
management strategies through a variety of intervention
structures that reflect the differing needs of parents.
The Triple P service modality is structured to enable parents
to access information from a variety of sources includingmultimedia, professional consultations and self-directed
modules. Distinct tiers of Triple P are available (http://www.
triplep.net/files/pdf/TripleP Model Tabole.pdf). These include
Standard, Group, Enhanced, Self-Directed, and Media
(i.e., one version of Universal Triple P). A child mean age of
five years was found in the Triple P evaluations included in
the current study.
Many of the published articles on the efficacy of Triple P
are based on Standard Triple P, Group Triple P or Enhanced
Triple P. Standard Triple P works with single families, while
group sessions are conducted within Group Triple P. Both
of these types of Triple P emphasize the role of parents in
the development and maintenance of child misbehavior by
assisting parents to identify possible causes and establishing
goals for behavior change (Turner, Markie-Dadds, &
Sanders, 1998). Through didactic presentations, individual
or small group activities and homework, parents use differential
reinforcement, communication skills, effective consequences
for misbehavior, and planned activity scheduling
(Turner et al., 1998). Approximately 10 sessions are available
for Standard Triple P, whereas, it is usual to offer five
group sessions and three telephone consultations to Group
Triple P participants.
In Enhanced Triple P, three adjunct modules, tailored
to the parents individual needs, are added to Standard
Triple P. Modules listed on the Triple P web site
(http://www1.triplep.net/) and in the Practitioners Manual
for Enhanced Triple P (Sanders, Markie-Dadds, & Turner,
1998) are Practice, Coping Skills, and Partner Support. In
the Practice module, the goals are to identify and resolve
problems with implementing new parenting strategies. In
the Coping Skills module, parents are assisted with personal
issues, such as depression and anxiety. In the Partner Support
module, dual parent families who are experiencing difficulties
in communication, relationships and/or co-parenting are
offered support (Sanders et al., 1998).
The final two interventions of Triple P are Self-Directed
and Media. Self-Directed Triple P participants are supplied
with a text and a self-help manual and provided with
10 weeks of structured learning tasks and do not have contact

with a professional (Sanders, Markie-Dadds, Tully, & Bor,

2000). Skills taught in Self-Directed Triple P are similar to
those taught in Standard Triple P and weekly telephone consultations
with a professional can be arranged. Media Triple
P is described as 12 episodes of an infotainment television
intervention (Sanders, Montgomery,&Brechman-Toussaint,
2000). Goals of the intervention are described as providing
parents with parenting strategies, to normalize the challenge
of parenting and to increase community awareness of family
relationships (Sanders, Montgomery et al., 2000).
Parent-Child Interaction Therapy (PCIT)
PCIT is an individualized intervention developed for caregivers
and their 4- to 7-year-old children with externalizing
behavior (Hembree-Kigin & McNeil, 1995). Therapeutic
outcomes of PCIT have been described as being guided
by observed changes in parent-child interactions, rather than
self-report changes in the childs behavior only. A parent is
assisted to alter her/his behavior via direct coaching strategies.
In turn, this change in parent behavior is expected to
improve the childs problem behaviors, and increase positive
interactions within the parent-child dyad. Although the
above is a shared principle of behavioral parent training, the
use of direct, in vivo coaching of parental behaviors differentiates
PCIT from most other behavior parent training
interventions, including Triple P. In the reviewed studies of
PCIT, the average length of treatment was between 12 and
14 weeks.
There are two phases in PCIT, which are labeled Child
Directed Interaction and Parent Directed Interaction. Progression
from one phase to the next is predominantly dependent
on attaining prescribed levels of specific skills known
as mastery criteria (Hembree-Kigin & McNeil, 1995). However,
in the studies reviewed here, PCIT was often limited
to a specific number of sessions (e.g., 12 sessions). PCIT
skills are taught via didactic presentations to parents, and direct
coaching of parents while they are interacting with their
children. In didactic sessions (usually two sessions), the focus
is on teaching the parent specific skills related to each
phase of the therapy and these sessions are conducted prior
to the direct coaching sessions. The remainder of PCIT (usually
about 1012 sessions in the studies reviewed) involves
direct coaching sessions. These sessions are conducted with
the parent and child in a play therapy room with the therapist
in another room behind a one-way mirror. The therapist and
the parent communicate through a bug-in-the-ear device.
This device permits the therapist to provide direct coaching
of parental communication and behavior management skills,
immediate feedback and social reinforcement of the parent.
Parents are expected to practice the skills at home.
As in Triple P, there are variants of the standard PCIT
intervention. Although PCIT was described as a 12-week intervention
in most studies, Nixon et al. (2003, 2004) used an
abbreviated version of PCIT that followed the same guidelines
as Standard PCIT (i.e., two phases of treatment), however
the two didactic sessions were replaced with videos
for home viewing. In addition, instead of the standard 10
12 PCIT direct coaching sessions, abbreviated PCIT offered
five in-vivo coaching sessions, which alternated with five
30-minute telephone consultations.

Another variant of PCIT adds a 6-session motivation

component (PCIT + motivation). This delivery format was
used in a randomized controlled trial of PCIT with families
who had a history of physically maltreating their children
(Chaffin et al., 2004). The motivation sessions included
videotaped testimonials from previous participants, psychoeducation,
decision making exercises, self-motivational
cognitions, self-efficacy, and an understanding of the
consequences of child maltreatment. Before proceeding to
Standard PCIT, participants submitted personal statements
regarding parental beliefs, practices and goals for therapy.
After PCIT, a 4-week group intervention was implemented
in order to improve generalization and maintenance of skills.
Also, Chaffin et al. (2004) implemented a second version of
Enhanced PCIT. In this version, participants were provided
with PCIT plus the motivational component, as well as
individual counseling sessions for depression, marital issues
or substance abuse, and home visits.
The delivery formats and specific intervention
strategies of Triple P and PCIT
According to Salas and Cannon-Bowers (2001), Triple P and
PCIT employ strategies that have been found to be effective
relevant information is presented to parents and parenting
concepts are demonstrated through role-play in both Triple
P and PCIT. Nevertheless, there are at least three key intervention
strategies that differ between PCIT and Triple P.
Summarizing the effects of these programs was expected
to provide important information about the relative efficacy
of different intervention components. First, a collaborative
learning model is inherent in Triple P group interventions
where group interactions assist in the learning process. Collaborative
learning has demonstrated efficacy and reduces
therapist time and resources (Shebilske, Jordan, Goettl, &
Paulus, 1998). PCIT does not have a group version. Second,
although parents in Triple P are encouraged to conduct homework
tasks and discuss these with the therapist, observed
practice of parents interacting with children and managing
challenging child behaviors does not usually occur as part of
the intervention design. Standard Individual and Enhanced
Triple P have the provisions for child involvement, but child
involvement was not described in the reviewed studies. In
contrast to Triple P, the use of coaching strategies that include
parents interacting with their children in PCIT is a key
intervention strategy and provides opportunities for parents
to practice new skills and get immediate feedback about performance
from interventionists. Because the parent might be
provided with direct remediation of incorrect implementation
of skills and practice with implementation that works
for the parent and the child, and because the overlearning
principles of PCIT evident in the Mastery Criteria allow
for repeated practice of skills and increases retention rates of
newly acquired skills (Driskell,Willis, & Copper, 1992), the
direct coaching strategies in PCIT might yield larger effects
on both parent and child outcomes. Third, although restricted
by a small number of published studies, comparing the effectiveness
of standard PCIT and Triple P to interventions
that have been enhanced with additional services and other
components was expected to add important information for
future intervention design and research.

The current study

In the current study, our aim was to provide a comprehensive
assessment of the state of the evidence to guide decisions
about the implementation and continued dissemination of
Triple P and PCIT in Australia and the USA, and to determine
whether summarizing the efficacy of PCIT and Triple
P allowed for conclusions about effective intervention delivery
format. To do this, we identified all randomized controlled
trials and single group follow-up studies of Triple
P and PCIT dated between 1980 and 2004. We abstracted
data from all identified studies, and used standard review
and meta-analysis techniques to draw conclusions about the
capacity of each type of intervention to improve parent and
child behaviors. We provided specific analyses of each subtype
of Triple P and PCIT that has been investigated and
summarized the efficacy of each type of intervention.We reported
effects sizes from independent group pre-treatment/
post-treatment (IGPP) studies and single group pre-treatment
/post-treatment (SGPP), and effect sizes based upon different
sources of information (e.g., parent-report, observation),
and compared effect sizes found in Standard PCIT to effect
sizes found for Triple P.
Method
Retrieval and selection of studies
We identified studies dated between 1980 and 2004 that
examined the efficacy of Triple P or PCIT. Five literature
search strategies were used. First, electronic databases
were searched (Medline and Psych INFO). Keywords used
to identify articles were Triple P, Positive Parenting Program,
Parent-Child Interaction Therapy and PCIT. In addition
to using the keyword function, searches were also conducted
on known Triple P and PCIT researchers (Eyberg,
Funderburk, McNeil, Nixon, Sanders and Turner). Second,
the reference sections of all retrieved articles were searched
for additional trials. Third, major reviews of parental behavioral
training and child behavior problems and reference
lists were inspected. Fourth, authors of identified trials were
contacted and all assisted in the identification of additional
publications. Fifth, our searches included book chapters and
attempts to obtain unpublished trials from known authors of
published trials of PCIT and Triple P, but no additional trials
were identified and retrieved.
Included and excluded studies
Studies reported between 1980 and 2004 were included in
the meta-analysis if they met three criteria: These criteria
included a focus on the efficacy of either Triple P or PCIT,
the inclusion of at least one parent or child behavior problems
outcomemeasure, and the inclusion of empirical data needed
for meta-analysis.
Thirty-two studies of PCIT were retrieved, and we included
13 studies from 8 cohorts and 3 research groups
(Chaffin 1 cohort, Eyberg/McNeil 6 cohorts, Nixon
1 cohort; see Table 1). Seven studies reported only means or
did not include needed information for meta-analysis, nine
studies were not effectiveness studies, one study considered
the effectiveness of PCIT within the classroom setting with
the teacher coached in PCIT skills (i.e., parents were not
included in the intervention), and one study was a duplicate
of results reported in an included study. 1 One further study

met most of the inclusion criteria, however, it was not included

because three treatment groups with different levels
of father involvement were compared and total effects for all
participants were not reported (Bagner & Eyberg, 2003).
We retrieved 28 studies of Triple P (or earlier versions
of this intervention). Eleven studies based on 10 cohorts
from 1 research group (Sanders) were included in metaanalyses
(see Table 1). Twelve studies did not include enough
information to be included in meta-analysis. Four studies were excluded as they focused
on child outcomes other than
child behavior problems (e.g., bed-wetting or pain).1 One
study metmost inclusion criteria, but results were reported by
subgroups in order to examinemartial discord as a moderator
of treatment effectiveness, and total results for all participants
were not reported (Dadds, Schwartz, & Sanders, 1987).
Data abstraction and coding
From each study, we abstracted study and intervention characteristics
(see Table 1). Each author abstracted information
independently and results were compared to correct inconsistencies.
All discrepancies were due to misreading article
content or data entry errors by one author. Hence, to correct
disagreements, the original reference was consulted and the
discrepancies were easily rectified. Intervention characteristics
that were extracted included the length of time in treatment,
the treatment components (e.g., additional individual
sessions for depression, psychoeducational information
on ADHD, marital relationship enhancement, etc.), and, for
Triple P, whether the interventionwas administered to groups
or individual parents/families. Other abstracted data were
characteristics of participants, study design, and outcomes.
When available, parent and child characteristics included
child gender, parent gender, race/ethnicity, and child age.
Study design factors included an indicator of randomization
versus matched/other comparison group, type of comparison
group, time in treatment, time to follow-up assessments, size
of treatment and comparison groups and attrition rate.
Outcome measures used in each study were classified as
negative or positive measures. With the exception of positive
and negative behavioral observations, scores on positive
measures were reversed so that higher scores always reflected
more negative parent and child behaviors.
PCIT study outcomes
The outcomes variables measured in studies of PCIT included
child problem behaviors as measured by parent- or
teacher-report, clinic or classroom observation, and parent
stress and behavior as measured by parent-report. Child problem
behavior was assessed with parent-report questionnaires
in all 13 PCIT studies, while teacher reports were gathered
in two published studies from a single cohort. The most
commonly used measures were the Eyberg Child Behavior
Inventory (ECBI, Eyberg & Pincus, 1999, mother report),
the Parenting Stress Inventory (PSI, Abidin, 1990, mother
report), and theDyadic Parent-Child Interaction Coding System
(DPICS, Robinson & Eyberg, 1981, observation).
We included more than one outcomemeasure from studies
of PCIT efficacy, M = 9 measurement subscale scores per
study with a range from 1 to 24 scores for each family; 10
studies from 6 cohorts included multiple measures of both child and parent outcomes,
while 3 studies from 2 cohorts

included only measures of child outcomes.

Tabel

child and parent outcomes, while 3 studies from 2 cohorts

included only measures of child outcomes.
Triple P study outcomes
The outcome measures in studies of Triple P included child
problem behaviors as measured by parent- or teacher-report
and clinic observation, while parenting stress and behavior
was measured by parent-report. Child and parent behaviors
were measured by parent-report questionnaires in all 11 studies
of Triple P. The most common parent-report measures
for child behavior outcomes included the ECBI and Parent
Daily Report (PDR, Chamberlain, & Reid, 1987). Clinic observation
of child behavior occurred in four studies using
a version of the Family Observation Schedule (FOS-R-III,
Sanders, Waugh, Tully, & Hynes, 1996, cited in Sanders,
Markie-Dadds et al., 2000). One study considered teacher
perceptions of child behavior problems measured by the
Sutter, Eyberg Student Behavior Inventory (SESBI; Eyberg
& Pincus, 1999).
We included more than one outcomemeasure from studies
of Triple P efficacy, M = 8 measurement subscale scores
per study with a range from 3 to 22 scores per family. Seven
studies from 6 cohorts included multiple measures of both
child and parent outcomes, while 4 studies from 4 cohorts
included only measures of child outcomes.
Descriptions of trial participants
In all included PCIT and Triple P trials, pre-treatment data
indicated that the average childwaswithin the clinical or borderline
range of externalizing behaviors, as measured by either
the ECBI (Intensity Scale), CBCL externalizing T score
or the Behavior Assessment System for Children (BASC)
externalizing T score. In 12 of the 13 studies of PCIT, the
mean score for child externalizing behavior was reported
within the clinical range. Seven of 11 studies of Triple P reported
pre-treatment child externalizing mean scores within
the clinical range, whereas the other four studies reported
pre-treatment means within the borderline range. Two studies,
one of PCIT (Chaffin et al., 2004) and one of Triple
P (Sanders et al., 2004), tested the efficacy of intervention
with families at risk of or engaged in maltreatment. Both
of these studies reported that the average child had a pretreatment
mean externalizing behavior score within the borderline
range.
Some studies restricted child participants by excluding
childrenwith developmental disorders and/or intellectual impairment
(Brestan, Eyberg, Boggs,&Algina, 1997; Connell,
Sanders, & Markie-Dadds, 1997; Eyberg et al., 2001; Hoath
& Sanders, 2002; Leung, Sanders, Leung, & Lau, 2003;
Sanders & McFarland, 2000; Schuhmann, Foote, Eyberg,
Boggs, & Algina, 1998), whereas exclusionary characteristics
of parents weremore varied. Parents intellectual impairment
was an exclusionary criterion in two studies of Triple
P (Hoath & Sanders 2002; Sanders, Markie-Dadds, Tully,
& Bor, 2000) and four studies of PCIT (Funderburk et al.,

1993; Hood & Eyberg, 2003; McNeil, Eyberg, Eisenstadt,

Newcomb, & Funderburk, 1991; Schuhmann et al., 1998).
In two other Triple P studies parents major psychiatric diagnosis
was an exclusionary criterion (Sanders & McFarland,
2000; Leung et al., 2003).
Some studies specified inclusionary criteria. In studies
of Triple P, these included parental depression (Sanders,
Markie-Dadds et al., 2000), marital conflict (Ireland,
Sanders, & Markie-Dadds, 2003; Sanders Markie-Dadds
et al., 2000), minimum 20 h/week employment (Martin
& Sanders, 2003), and parental notification of child maltreatment
(not necessarily confirmed) or elevated scores
on three subscales of the State-Trait Anger Expression Inventory
(Sanders et al., 2004). One study of PCIT specified
the requirements of a confirmed notification of child
physical maltreatment (Chaffin et al., 2004). The other
12 PCIT studies did not specify inclusionary criteria for
parents.
Computation and analyses of effect sizes
We computed single group effects from pre-treatment
to post-treatment (SGPP) and pre-treatment to followup
(4 months to one year after treatment completion;
SGPF).We also analyzed independent groups pre-treatment/
post-treatment (IGPP) and independent groups pretreatment/
follow-up (IGPF) effects. In the following sections,
we refer to treatment versus comparison groups in
the IGPP and IGPF analyses. Comparison groups included
randomized waitlists, matched control groups, and alternative
community intervention or social validation (normal
community) groups (see Table 1).
SGPP and SGPF effect sizes
The formula used to calculate a single group effect size
(SGPP or SGPF) was the following:
dsg = (Mpost Mpre)/SDpre
where,
dsg was the single group effect size (SGPP or SGPF) reported
in Tables,
Mpost was the mean value at post-treatment or at follow-up,
Mpre was the mean value at pre-treatment, and,
SDpre was the standard deviation at pre-treatment.
In one single group study, means and standard deviations
were not reported, so paired t-test values were used to calculate
effect sizes.
IGPP and IGPF effect sizes
As suggested by Becker (1988) and Morris & DeShon
(2002), the final IGPP and IGPF effect sizes used in analyses
were calculated as the difference between the treatment
and comparison single group effects. The three formulas
used to calculate IGPP effect sizes and IGPF effect sizes
were:
dt = (Mt-post Mt- pre)/SDt-pre
dc = (Mc-post Mc-pre)/SDc-pre
dig = dt -dc
where,
dt was the effect size for the treatment group,
dc was the effect size for the comparison group,
dig was the final IGPP or IGPF effect size reported in the
Tables and Figures,

Mt-post and Mc-post were the mean values for the treatment
group (t-post) or the comparison group (c-post) at posttest
or at follow-up,
Mt-pre and Mc-pre were the mean values for the treatment
group (t-pre) or the comparison group (c-pre) at pre-test,
SDc-pre and SDc-pre were the standard deviations at pretest.
Hence, dig was the treatment effect size net of the comparison
group effect size. When dig was positive this indicated
a larger effect in the treatment group than in the comparison
group. When dig was negative this indicated a larger effect
in the comparison group than in the treatment group.
We used pre-test SDs for all calculations, as suggested
by Becker (1988), because pre-test SDs are often more
consistent across studies than post-test SDs. The use of pretest
SDs to estimate effect sizes, therefore, results in estimates
that are more comparable across different experimental
manipulations (in this case, between studies of PCIT and
Triple P). Post-test SDs have been found to be less consistent
than pre-test SDs as a result of different experimental
manipulations. We kept single group effect sizes separate
from independent group effect sizes, because there is continuing
debate about when these can be combined, and because
single group effect sizes are often larger than independent
group effect sizes (see Carlson & Schmidt, 1999; Morris &
DeShon, 2002). Although there are many reasons that single
group and independent group effect sizes might differ and
both study designs can have their own biases, one reason
single group effect sizes often are larger is because they do
not take into account the time effect (e.g., history or maturation
effect) that can be partly accounted for by a comparison
group (Morris & DeShon, 2002).
PCIT effect sizes
Thirteen studies were included in calculations of SGPP effect
sizes; eight studies were included in calculations of IGPP
effect sizes. Four studies were available for calculations of
SGPF and IGPF effects.We calculated the following for child
behavioral outcomes: 65 SGPP effects, 30 SGPF (4 month to
1 year after treatment) effects, 49 IGPP effect sizes, and 14
IGPF effect sizes. For parenting behaviors, we calculated 47
SGPP effects, 14 SGPF (4 month to 1 year after treatment)
effects, 54 IGPP effects, and 16 IGPF effects.
Triple P effect sizes
Eleven studies were included in the calculations of SGPP;
whereas seven studies were included in calculations of IGPP
There were no studies with comparison groups that included
a follow-up assessment.We calculated the following for child
behavioral outcomes: 78 SGPP effects, 27 SGPF (4 month to
1 year after treatment) effects, and 53 IGPP effects. For parenting
behaviors, we calculated 77 SGPP effects, 55 SGPF (4
month to 1 year after treatment) effects, and 50 IGPP effects.
Reduction of effect sizes and final analysis
Effect sizes were categorized by purpose of the measure (to
assess child or parent), method (questionnaire versus observation)
and reporter (mother, father, teacher). In some cases,
effect sizes in the same category within a study were averaged
after accounting for repetition from multiple studies
from the same sample. This averaging was done to reduce
bias that might be introduced from studies that used more
measures, and findings reported in more than one publication

were included only once. In sum, we report one effect

size per study sample for (a) child behavior by method and
reporter and (b) parenting by method and reporter. Effect
sizes were analyzed with DSTAT (Johnson, 1989) to compute
bias-corrected summary effect sizes, confidence intervals,
r values, to examine homogeneity of effect sizes and
outliers, and to compare effect sizes for PCIT to those for
Triple P. Since there was homogeneity of effect sizes in all
analyses, we did not examine correlates of effect sizes.
Most observational and survey measures used within trials
of PCIT and Triple P were for the assessment of negative
child or negative parenting behaviors. Hence, except
in the cases of observations of positive child or parent behavior,
negative effect sizes were expected as they would
indicate declines in problem behaviors from pre-treatment
to later assessments in studies with single group repeated measurement designs, and
greater declines in the treatment
group compared to the comparison group in studies with independent
groups repeated measurement designs. With the
exception of one PCIT study (Eyberg & Robinson, 1982),
coding of observational measures in all studies was reported
to have been conducted by coders who were independent of
the study and blind to participant status.
Results: PCIT
Child behavior change during treatment: Single group
treatment effects
In PCIT, medium to large effect sizes were observed for single
group child behavior change from pre-to post-treatment
(see Table 2). Improvements were found in clinic observations
of both negative and positive child behavior, d = .54
and .94, respectively, and large effects were found when outcomes
were based on mother and father reports of child
negative behavior, d = .1.31 and .83, respectively. In
one study of 10 children, no significant changes were found
in classroom observations of positive child behavior and
teacher report of negative child behavior from pre to post
treatment.
Effect sizes for child behavior change from pre-to followup,
however, were less likely to be significantly different
from 0. The only statistically significant effect size was for
mother (or mixed mother/father) report of negative child
behavior, d = 1.10. No significant change in child behaviors
from pre-treatment to follow-up was found when effect
sizes were based on clinic observation, father report (when
isolated from mother report), teacher report, or classroom
observation.
Child behavior: Treatment versus comparison groups
Standard PCIT
When compared towaitlist, medium and large effects ranging
from absolute values of .611.45 were found favoring PCIT
formother (ormixed mother/father) reports of negative child
behavior, and father reported negative child behavior (see
Table 2). However, no significant effect was found for observed
negative child behaviors. When PCIT children were
compared to normal (i.e., nonclinical) comparison groups,
mothers in PCIT reported greater declines in their childrens
behavior problems than other mothers and large effects were
also found for positive behaviors as observed in the classroom,
and teacher reports of negative behaviors of children

in PCIT compared to a normal comparison group. Effect

sizes ranged from an absolute value of 1.211.57. When
compared to a deviant community group, teachers reported
greater improvements in negative behaviors of children in
PCIT d = 1.16 however there were no significant effects
for observations of positive behavior in the classroom.
Abbreviated PCIT
No significant effects of Abbreviated PCIT were found
when child behaviors were compared to waitlist comparison
groups (see Table 2). When Abbreviated PCIT participants
were compared to a normal community comparison group,
there was no significant effect on observed positive child
behavior, but there was significant improvement for PCIT
participants when child behaviors were assessedwith mother
report, d = 1.57.
Enhanced PCIT
No comparisons of Enhanced PCIT (PCIT + motivation or
PCIT + motivation + individual) versus waitlist comparison
groups were conducted. Instead, in one study, parents
with a history of maltreating their children were assigned to
PCIT + motivation or a community group didactic intervention.
In this study, parent-reports of child behavior were
collected, and the effect of PCIT was large, d = .83 (see
Table 2). A large effect on mother-reported child behavior
problems also was found for a PCIT + motivation + individual
intervention when compared to a community group
didactic intervention, d = 2.16.
Parent change during treatment: Single group
treatment effects
There were significant changes in parenting outcomes pretreatment
to post-treatment,with effect sizes for negative and
positive parenting ranging from an absolute value of 1.11
3.11 (see Table 3). The only exception was a nonsignificant
effect size found for father-reported negative behaviors in
one small study of 12 fathers, d = .68.
Medium to large effectswere found for clinic observations
of negative parent behaviors and parent-report measures of
negative parent factors from pre-treatment to follow-up. Effect
sizes ranged from an absolute value of .61 to .94.
Parenting: Treatment versus comparison groups
Standard PCIT
There was more improvement in parent behavior and functioning
among parents in PCIT than those in a waitlist. All
effect sizes except one (observations of fathers negative
behaviors in one small study of 22 fathers) were significant
and usually large in magnitude, d ranged from an absolute
value of .76 to 5.67 (see Table 3). Similar large effects of
PCIT were found for observations of mothers, d = 1.03,
and mother-reports of their parenting, d = 1.59, when
compared to mothers in a normal community comparison
group.
Tabelll
Abbreviated PCIT
Compared towaitlist, medium or large effects ofAbbreviated
PCIT were found for observed changes in positive, d = .92,
but not negative, parenting behavior and mother-reports of

their parenting, d = .74 (see Table 3). When negative

parenting behaviors of participants in Abbreviated PCIT
were compared to a normal community group, there were
moderate to large effects on clinic observation of negative
parental behavior, d = .82, and mother reports of their
behaviors, d = .75.
tabel
Enhanced PCIT
There was a large effect when Enhanced PCIT + motivation
was compared to a community group didactic intervention
favoring Enhanced PCIT on all parent measures,
d ranged from an absolute value of 1.15 to 1.65 (see
Table 3). Enhanced PCIT + motivation + individual also
was compared to a community group didactic intervention
on all parent measures. Large effect sizes ranging from an
absolute value of .86 to 4.79 favoring Enhanced PCIT +
motivation + individual also were found for observed negative
and positive parent behavior and negative parenting
based on self-report.
Results: Triple P
Child behavior change during treatment: Single group
treatment effects
For Triple P, small and medium effects were found for
single group pre- to post-treatment child behavior (see
Table 4). Clinic observed negative child behavior, negative
child behavior measured by both mother and father reports
and stepparent reports of negative child behavior improved.
Effect sizes ranged from an absolute value of .31 to .73.
However, a small study of 9 children with teacher reports of child behavior did not yield a
significant effect size. Analyses
of single group pre-to follow-up data on child behavior
resulted in small and medium effect sizes for all measures, d
ranged from an absolute value of .36 to .70 (see Table 4).
Tabell
Child behavior: Treatment versus comparison groups
Standard Individual Triple P
Medium effects were found in favor of Standard Individual
Triple P when compared to waitlist comparison group when
child negative behavior was reported by mothers (or mixed
mothers/fathers), d = .69 or by fathers, d = .60 (see
Table 4). This did not hold for clinic observations of negative
child behavior and one study including reports from 42
stepparents.
Group Triple P
When compared to a waitlist control, a medium effect was
found for Standard Group Triple P when based on mother
(or mixed mother/father) report, d = .67 (see Table 4).
No studies used observational methods or collected reports
from individuals other than parents.
Enhanced Triple P
When Enhanced Triple P was compared to waitlist, medium
and large effect sizes favoring Enhanced Triple P were found
for clinic observations of negative child behavior and father

and mother reports of negative child behavior. Effect sizes

ranged from an absolute value of .46 to .96 (see Table 4). In
a small study of 20 teachers, there was no significant effect of Triple P when negative
child behavior was compared to a
waitlist comparison group.
Self-directed Triple P
Mothers in Self-Directed Triple P reported moderate improvements,
d = .51, while fathers reported large improvements
in their childrens behavior, d = 1.26, compared
to waitlistmothers and fathers (see Table 4). No significant
effect on clinic observations of child negative behavior
was found.
Media Triple P
Compared to waitlist, Media Triple P had a large effect on
child behaviors as reported by parents, d = .79.
Parent change during treatment: Single group
treatment effects
Small and medium effects on parenting behaviors were
found in analyses of single group pre-to post-treatment data,
d = .70 for mother report and d = .38 for father report
(see Table 5). Therewas no significant effect for observations
of negative parent behavior. Small and medium effects were
found for pre-to follow-up changes in parenting self-report
measures for mother report and father report and observed
negative parent behavior ranging from an absolute value of
.28 to .69.
Parenting: Treatment versus comparison groups
Standard Individual Triple P
When compared to waitlist, Standard Individual Triple P
had a large effect favoring the Triple P intervention when
based on mother (or mixed mother/father) reports of parenting,
d = 1.07, and a medium effect based on father
reports of parenting, d = .40 (see Table 5). However,
this did not hold for clinic observed negative parent
behaviors.
Group Triple P
Compared to a waitlist comparison group, a medium effect
size was found in favor of Standard Group Triple P when based on mother (or mixed
mother/father) reports of their
parenting, d = .69 (see Table 5).
Tabell
Enhanced Triple P
When comparing Enhanced Triple Pwithwaitlist, therewere
statistically significant effects favoring Enhanced Triple P
for measures of negative parent factors as reported by mothers
(or mixed mother/father; large effect), d = .98 and
fathers, d = .46 (medium effect, see Table 5). This
did not hold for clinic observations of negative parent
behaviors.
Self-directed Triple P
When we analyzed findings comparing Self-Directed Triple
P to waitlist, large and small effects favoring SelfDirected Triple P were found based on mother (or mixed
mother/father) report, d = 1.44, and father report,
d = .36, respectively (see Table 5). This did not hold
for observed negative parent behavior.

Media
There was no significant effect of Media Triple P on parenting
based on self-report questionnaire.
Results: Comparison of PCIT and Triple P
We used categorical modeling to compare the effect sizes
for PCIT and Triple P (Johnson, 1989). We conducted analyses
to compare SGPP effect sizes, and compared IGPP for
Standard PCIT (compared to waitlist) to the multiple forms
of Triple P (compared to waitlist).
Single group pre-treatment/post-treatment (SGPP)
effect sizes
For SGPP, PCIT had larger effects of parent report of child
negative behaviors, but not when effect sizes were based
on observed child behaviors. More specifically, there was
a larger improvement for study families in PCIT than in
Triple P for parent-report of child behavior problems, 1.31
vs. .73, p<.001 (see Tables 25). However, there was no
difference in SGPP effect size for observed child negative
behavior, .54 vs. .31, p = .27.
Findings when comparing SGPP for parenting problems
showed that PCIT had larger effects for both parent report
and observed parent negative behaviors. There was a larger
improvement for study families in PCIT than in Triple P
when outcomes were based on parent report of parenting
problems, 1.11 vs. .70, p<.001. There also was a larger
improvement in observed parent negative behavior for PCIT
than Triple P, 1.46 vs. .19, p<.001.
Independent groups pre-treatment/post-treatment
(IGPP) effect sizes
Child negative behaviors
Figure 1 shows IGPP effect sizes and confidence intervals
for the four forms of PCIT and five forms of Triple P when
compared to waitlist. The effects in Fig. 1 are based on
parent-reports of child negative behavior. As shown, there
were large effect sizes for child negative behavior in all forms
of PCIT, except the Abbreviated version, which did not have
a significant effect on child negative behavior. There were
medium to large effects for all forms of Triple P.
As can be seen by comparing the confidence intervals in
Fig. 1, results of our analyses that compared Standard PCIT
to multiple forms of Triple P showed differences in PCIT and
some forms of Triple P for parent report of child negative
behaviors. The effect size for PCIT, 1.45,was significantly
larger than the effect sizes for Self-directed, .51, p<.001,
Group, .67, p<.01, and Individual Triple P, .69, p<.01,
but the PCIT effect size was not larger than the effect size
for Enhanced, .96, or Media Triple P, .79. However, in
contrast to these findings, for IGPP based on observed child
negative behaviors, there were no differences in the effect
size for Standard PCIT, .11, compared to Triple P in the
Self-directed, .02, Individual, .22, and Enhanced, .46,
forms (figure not provided).
Parent negative behaviors
Figure 2 summarizes IGPP effect sizes based on parent reports
of parent negative behavior, and illustrates that all forms
of PCIT and Triple P, except Media Triple P, had medium or
large effects. Media Triple P did not have a significant effect
on self-reported negative parent behavior. All other interventions
reduced negative parenting of those in treatment

when compared to waitlist. Of these effective interventions,

Abbreviated PCIT and Group Triple P had the most modest
effects on parenting, but effect sizes were still medium in size
and quite similar to other, often more intensive and longer
forms of these interventions (i.e., all confidence intervals had
a range of values in common).
When we statistically compared Standard PCIT to forms
of Triple P, PCIT had a larger effect when compared to
Media Triple P, but not when compared to other forms of
Triple P; the effect size for Standard PCIT, 1.16, was not
different when compared to Self-directed, 1.44, Group,
.69, Individual, 1.07, and Enhanced Triple P, .98. Yet,
PCIT did have a larger effect on parent report of parenting
problems than media Triple P, .45, p<.05. However, for IGPP of observed parent
negative behaviors, Standard PCIT
had a larger effect size, .76, than Triple P in the Selfdirected,
.07, p<.01, Individual, .14, p<.01, and Enhanced,
.11, p<.05, forms (figure not provided).
Gambarrr
Discussion
Behavioral parent training is a popular form of parenting
intervention (McMahon & Forehand 2003; Nixon, 2002;
Sanders et al., 2003; Webster-Stratton & Taylor 2001)
and has been described as efficacious in previous metaanalytic
and reviewarticles (Barlow&Stewart-Brown, 2000;
Serketich & Dumas 1996). In an attempt to go beyond statements
pertaining to the efficacy of behavioral parent training
as a broad category, our review and meta-analyses described
the reported efficacy of two parenting interventions with
known behavioral parent training origins. The interventions
were chosen for this review due to their similar theoretical
foundations, and their wide-spread dissemination, popularity
and significant level of government funding in both the
USA and Australia, but their differing modes of delivery.
Gambarrr
We analyzed 24 evaluations of Parent-Child Interaction
Therapy (PCIT) and Triple P-Positive Parenting Program.
Findings revealed that these interventions improve parenting,
such as improving parental warmth, decreasing parental hostility,
increasing parental self-efficacy, and reducing parental
stress. Most versions of these interventions also reduce negative
child behaviors, such as aggression and extreme tantrums
and opposition. These results put numerical weight behind
previous narrative reviews in which authors have concluded
that interventions are effective when they include behavioral
parent training (Nixon, 2002; Webster-Stratton & Taylor,
2001; Weisz, Hawley, & Jensen Doss, 2004).
Participation in PCIT and Triple P results in improvements
in child behavior and parenting from pre- to post-treatment
(i.e., short-term improvements in behavior). In addition,
although few studies included follow-up beyond immediate
post-treatment, there was some support from prospective
assessment of treatment participants for continued positive
effects of these interventions up to 3 months after intervention
completion. However, conclusions regarding long-term

effectiveness of either intervention must remain tentative.

Follow-up data collected from treatment participants were
never compared to waitlist or alternative treatment groups.
More specifically, follow-up data were not collected from
waitlist participants for any of the Triple P studies and followup
data only were collected from a social validation group
in only one series of PCIT studies in Australia (Nixon et al.,
2003, 2004).
Findings show that the strength of the effects of PCIT
and Triple P, and conclusions about which intervention may
result in greater improvements in parenting and family functioning,
depended on the measures used to assess outcomes
and the subtype of each intervention. Although findings were
not without exception, Standard PCIT tended to have larger
effects than Triple P when compared to waitlist and when
outcomes were based on parent report of child negative behaviors
and observed parent negative behaviors. In contrast,
there was no effect size difference when findings were based
on observed child behaviors and only one difference (Standard
PCIT vs. Media Triple P) for parent report of parenting.
In addition, Standard PCIT did not have a larger effect than
Enhanced Triple P, except when comparing observed parent
negative behavior.
For some intervention types and when observational measures
were used, effects were not consistently found to be
different from 0 (i.e., effectswere not significant). Significant
treatment effects on childrens behaviors were found for both
Triple P and PCIT interventions when outcomes were assessed
via female caregiver (or combined female/male caregiver)
reports or observation. Effects were generally medium
for Triple P and large for PCIT. The exceptions were a large
effect size for Enhanced Triple P and a small and nonsignificant
effect size for Abbreviated PCIT. Effect sizes based
on teacher reports were calculated on single studies with
small samples, and were rarely as large as those based on
female caregiver report or observation and often did not
show significant improvements in childrens behaviors, but
improvements among PCIT participants were greater than
waitlist comparison groups (Funderburk et al., 1998; Hoath
& Sanders, 2002; McNeil et al., 1991). Large effect sizes
were found for father reports of child behavior and parenting
when PCIT was compared to a waitlist (Brestan et al.,
1997; Nixon et al., 2003, 2004; Schuhmann et al., 1998) and
small to large effects were found with the same measures
for Enhanced Triple P and Self-Directed Triple P (Connell
et al., 1997; Sanders, Markie-Dadds et al., 2000). Parenting
during Self-Directed Triple P did not significantly improve
when father reported behaviors.
Considerations for the generalization of study findings
There are three key issues to consider when generalizing the
results of the current meta-analyses. All issues come from
the sampling details included in studies and the measurement
strategies used. First, the demographic characteristics
of families included in many of the studies were unclear or
may have been limited to moderate or higher income families.
Only two of the Triple P studies had participants within
the lower socio-economic status (SES) group and with low
parent education (Hoath & Sanders, 2002; Sanders et al.,
2004). Sanders and McFarland (2000) reported participants

in the low SES range, however failed to report parental education.

Although sometimes unclear, all other participants in
Triple P studies appeared to have been from middle class or
higher SES groups and, with the exception of one study (Bor,
Sanders, & Markie-Dadds, 2002), middle or higher levels of
parental education. Due to the high number of Triple P studies
in the meta-analysis with middle or higher SES, it is not
certain that findings can be generalized to low income or
high risk groups at this time.
With the exception of three studies from one study group
(Nixon, 2001; Nixon et al., 2003; Nixon et al., 2004), participant
demographic data were not consistently reported
in studies of PCIT. When income or educational informationwas
reported, participantswere low-to-mid SES (Eyberg
et al., 2001;Hood&Eyberg, 2003;Nixon, 2001;Nixon et al.,
2003, 2004; McNeil et al., 1991; Schuhmann et al., 1998).
In the cases when education was reported, PCIT participants
had moderate levels of education (Brestan et al., 1997;
Eyberg & Robinson, 1982; Nixon, 2001; Nixon et al., 2003,
2004). Given the inconsistency in reporting of demographic
data, it also is difficult to draw conclusions about generalization
of PCIT to low SES and high-risk families.
Second, themethods of recruiting families differed. Triple
P studies most often used media outlets to advertise and recruit
families via self-referral (9 of the 11 Triple P studies
used media as a recruitment strategy), while PCIT participants
were most often clinic-referred (only 3 of the 13 PCIT
studies used media as a recruitment strategy).Only oneTriple
P study used clinic referred participants (Leung et al., 2003).
It remains unclear howmuch of the difference between PCIT
effect sizes and Triple P effect sizes may be due to differing
recruitment methods. Previous researchers have suggested
that there will be selection bias effects when participants
are recruited via the media and self-referred as compared
to relying on referrals from clinics or other professional
sources (Berlin, ONeal, & Brooks-Gunn, 1998). Reyno and
McGrath (2006) found a moderate effect size when comparing
source of referral and treatment outcome in favor of
self-referral over referral from professionals. However, we
found no evidence of this in the current study. Although,
there was some variability in recruitment strategies across
PCIT studies, and across Triple P studies, homogeneity of
effect sizes was found within studies of PCIT and within
studies of Triple P. This suggests that recruitment methods
may not have greatly influenced effect size.
Third, parent-reported child behavior problems were assessed
with either the Eyberg Child Behavior Inventory
(Eyberg & Pincus, 1999) or the Child Behavior Checklist
(Achenbach, 1991) in all but one study included in this review.
Of studies of PCIT, the ECBI was used in nine studies,
the ECBI and the CBCL were used in three studies, and the
BASC was used in one study. Similarly, in studies of Triple
P, the ECBI was used in nine studies, and the CBCL was
used in two studies This resulted in more meaningful comparisons
of the effect sizes for childrens behaviors found in
studies of PCIT to those found in studies of Triple P.
We also compared intervention types on effect sizes for
observed child and parent behaviors and parent-reports of
their own parenting behaviors. In contrast to parent reports

of child behaviors, there were some differences in the observational

measures of children and parents, and measures
that assessed parenting, and we cannot rule out the possibility
that differences are due to the use of different measures.
For example, DPICS (Robinson & Eyberg, 1981) was the
observational coding system used in studies of PCIT. DPICS
coding resembles and follows directly from treatment activity
(i.e., parents are coached in specific communication skills
such as praises, descriptions and reflections and DPICS is an
observed frequency count of these verbalizations). Progression
through treatment often depends on seeing improvements
in DPICS scores. In Triple P, a different observational
coding system (the Family Observation Schedule, Sanders et
al., 1996, cited in Sanders, Markie-Dadds et al., 2000) was
used to assess observed behaviors of children and parents.
Although the FOS was developed to assess the aims of Triple
P it does not directly follow treatment activity nor is progression
of treatment relative to this measure, whereas coaching
of PCIT skills is directly linked to the DPICS. Further, families
are repeatedly videotaped and coded for observed behaviors
(using DPICS results throughout PCIT sessions to
guide intervention strategies) and this is not done in Triple
P. These differences may have made it more likely to find
observed improvements in child and parent behaviors among
families in PCIT (Weisz, Weiss, Han, Granger, & Morton,
1995). Overall, including observational measures in addition
to parent report measures is recommended for future studies,
and the finding of more significant effect sizes in PCIT than
in Triple P when using observational measures may show
that it is advantageous, both for intervention development
and outcome studies, to develop observational measures that
are designed to test specific intervention aims (as has been
done in PCIT).
Implications for clinical practice and intervention
dissemination
Triple P and PCIT both have been designed to provide participants
with a range of intervention options. For example,
therapist contact time and length of service delivery varies
for Triple P - from Media Triple P and Self-Directed Triple
P with minimal therapist involvement, to Group Triple P
and Standard Individual Triple P. Enhanced Triple P also
is available, which is administered individually and includes
additional sessions on coping skills and partner support. Due
to the limited number of comparison studies of various intervention
types, it is difficult to determine the benefits of
the enhanced intervention versions. Currently the findings
do not clearly support the additional benefits of concurrent
treatment options for families, such as those provided in Enhanced
Triple P and Enhanced PCIT. When compared to a
waitlist control, Self-Directed Triple P (with minimal therapist
contact) produced similar effect sizes when compared to
Enhanced Triple P. Larger effects were found for PCIT with
the addition of a motivational enhancement and individual
services (PCIT+ME+ I) when compared to Abbreviated
PCIT. However, these large effect sizes for PCIT+ME+I
were likely influenced by the structure of the intervention.
Chaffin et al. (2004) reported that participants were required
to succeed in the motivational component prior to commencing
PCIT. This requirement may have inflated effect sizes,

because participants who entered and completed PCIT had

already completed a 6-week motivational course and, although
retention rates for the motivation course were reported,
the overall attrition rate was unclear. These mixed
findings are consistent with Reyno and McGraths (2006) assertion
that further study is required to determine whether the
addition of concurrent treatment in interventions,which have
already been shown to be efficacious, provides any additional
positive outcomes for participants. It is possible that participants
may not be able to maintain commitment to a variety
of intervention approaches simultaneously, and treatment
providersmay not have the resources to develop and maintain
high standards across all intervention components.
Although PCIT and Triple P have overlapping content, a
further implication for clinical practice may be embedded in
the differences in design and intensity of each intervention.
The somewhat larger, and sometimes significantly larger, effect
sizes for Standard PCITwhen considering child behavior
change may illustrate that intensity and design matter to the
level of positive outcomes for parents and children. With
respect to child behavior change, it appears that the direct
involvement of both the parent and the child in joint therapeutic
sessions that provide parents opportunities to practice
new skills may produce somewhat larger improvements, especially
in parent reports of child behavior problems and
observed parent negative behavior. However, there was less
evidence that there were greater improvements in observed
child negative behavior in PCIT than in Triple P, and parents
did not tend to report greater improvements in PCIT than in
Triple P. In sum, a more firm determination of differences in
behavior change and parenting between interventions awaits
further studies to add to the meta-analyses conducted here.
Nevertheless, these findings provide some initial support for
the usefulness of the intervention methodology of PCIT of
enhanced parental practice of skills under direct observation
and coaching by a therapist.
Of further interest to clinicians is the notion of dissemination
and transportability of interventions, which are designed
and tested in controlled clinic environments. Both PCIT and
Triple P have been disseminated to community agencies in
the USA and Australia, however, there have been no published,
independent dissemination or transportability studies
from either intervention. This means there is no current evidence
for effectiveness of either Triple P or PCIT in a community
setting. Triple P and PCIT could enhance evaluations
of intervention effectiveness by evaluating the respective interventions
in a diversity of settings. As dissemination of
each intervention continues, independent evaluations by the
community services now implementing PCIT and/or Triple
Pwould increase the knowledge of intervention effectiveness
and the effects of dissemination.
Conclusions
According to the current guidelines for evidence-based practice
in psychology, which are limited in a number of ways
by not including criteria for length of follow-up or study
sample size (American Psychological Association, 1995;
Chambless & Hollon, 1998), we tentatively conclude that
PCIT meets the criteria for a well-established treatment
and Triple P meets the criteria for a probably efficacious

treatment.We were precluded from describing Triple P as a

well-established treatment, because our search revealed that
Triple P evaluations have not yet been conducted by two
independent investigators or investigatory teams. Given the
need to make choices about how resources are allocated for a
range of interventions, independent trials of PCIT and Triple
Pwith longer-term follow-up and cost-effectiveness analyses
should be conducted.
Researchers are charged with the responsibility to provide
the community with information that assists effective
decision-making. Multiple trials of both PCIT and Triple P
have greatly assisted in developing two parenting interventions
to assist parents to have better relationships with their
children and to reduce child behavior problems. This provides
the community with much needed resources. Some of
the remaining critical future questions arewhich intervention
and what components are most beneficial for which population
and in what context. To determine this, further studies
specifically examining participant demographics, comparisons
to alternative treatment options and independent
evaluations of Triple P will assist clinicians in determining
which intervention would best meet the needs of their client
and in doing so could possibly alleviate future years of family
dysfunction and parenting stress, and alter pathways of
delinquent and antisocial behavior among a large number of
children and adolescents.
Acknowledgements This project was greatly improved through our
conversations with honours and postgraduate students working in the
Family Interaction Research Program at Griffith University, including
Elbina Avdagic, Angela Anthonysamy, Michelle Hanisch, Kate
McCarthy, Leanne McGregor, Rhiarne Pronk, Mark Scholes, Anne
Stuksrud, and Judith Warner. We also thank members of the Griffith
Psychological Health Research Centre at Griffith University for their
willingness to discuss some of the ideas presented in this article.